Your Book Review: Humankind

Finalist #12 of the Book Review Contest

[This is the twelfth of many finalists in the book review contest. It’s not by me - it’s by an ACX reader who will remain anonymous until after voting is done, to prevent their identity from influencing your decisions. I’ll be posting about two of these a week for several months. When you’ve read all of them, I’ll ask you to vote for your favorite, so remember which ones you liked. If you like reading these reviews, check out point 3 here for a way you can help move the contest forward by reading lots more of them - SA]

Human nature is usually said to be basically selfish and sinful, but Rutger Bregman begs to differ. In Humankind he argues that human nature is basically kind and decent. Unfortunately, his approach seems to have been inspired by Monty Python: in the introduction he builds a sparkling argument, then in section one he accidentally sets it on fire, knocks it over, and then watches it sink into the swamp. Then in section two he rebuilds it, only to douse it in petrol, and then leave the chip pan on in section three. By the end of this review we'll have unearthed some important truths. None of them will be "we can trust Bregman for logical consistency and factual accuracy".

Introduction - Good arguments that crises bring out the best in people
If at first you don't succeed, call in an airstrike.
Before the Blitz the consensus was that a little light bombing was all it took to make the wheels come off civilisation. This is based on veneer theory - our good behaviour is a thin veneer laid on our fundamentally selfish, violent nature, and that under pressure our true nature will out.

This turned out not to be true. So spectacularly untrue that we still talk about the Blitz Spirit. With our trademark humility, the British concluded that this was due to our exceptional moral fibre and, with help from the Americans, set about bombing German civilians to hell and back. Regrettably the Germans too responded by pulling together, and working harder in the war effort. Literally no one thinks this was due to their exceptional moral fibre. Instead, it seemed that crisis led to teamwork. Bregman is able to quote similar behaviour on the Titanic, on September 11th and in the aftermath of Hurricane Katrina.

Despite this mountain of evidence, veneer theory is still overwhelmingly believed. In 1951 William Golding wrote Lord of the Flies - a book about how a group of British boys crash-landed on a Pacific island would really behave. They start with ideals of co-operation, but quickly descend to violence and anarchy. Weeks later when they're rescued half of them are dead. The book became a massive best seller, and a much-studied classic. For those who lived through World War I, World War II, and were now watching communism demonstrate that you didn't even need an enemy to slaughter tens of millions, you can see the appeal of a cynical view of human nature. However it is pure fiction. In 1966 Lord of the Flies happened for real - 6 teenagers went for a joy ride in a fishing boat, got swept out by a storm and washed up on an inhospitable island in the Pacific. When they were found 11 months later, they were all alive and healthy. They had survived by fortitude, resourcefulness and above all, teamwork.

If you think people are screwed up, you will screw up
You can do surveys asking people how they will behave in certain situations, and how they think people in general will behave, and the answers are very consistent: people say they will behave well, as will the people they know well, but they expect people in general to behave badly. When shown people behaving altruistically subjects assume they have ulterior motives. When shown data about how often humans are altruistic, they come up with increasingly elaborate theories about how the behaviour is cynical really. "Cynicism is a theory of everything" writes Bregman. We live in a world of people who pull together in a crisis, but we believe we live in a world where people turn nasty in a crisis. Bregman blames the media for this (but in case that wasn't original enough on the next page he will blame scientists and religion) - the news serves us up the sensational and appalling, and because it serves it up every day it's easy to mistake it for the representative. He goes on to share studies that find watching the news is addictive and bad for you (at least, that's my excuse next time I'm found ignorant of current affairs). 'Reality TV' turns out to involve massive manipulation to get the contestants to be mean to each other and generate some interesting television (so don't give up on cynicism entirely, I guess?). I don't know of any studies showing Reality TV is bad for your mental health, but I'm happy to take it as read. This matters, because what we think people are like affects our choices and our behaviour. Bregman has a term for this - a 'nocebo'. In 1999 a mystery illness in 9 Belgium schoolkids who had drunk Coca Cola led to the suspicion that a batch had been accidentally poisoned. It was reported in the press, and Coke recalled 17 million cases, but it was too late - symptoms spread through Belgium and into France. A few weeks later toxicologists issued their report - there was absolutely nothing wrong with the coke (well, it's frighteningly high in sugar, and will dissolve your teeth, but there was nothing unusually wrong with the coke). Just as placebos can make us feel better even though in truth they're nothing, nocebos can make us feel ill, even though they're not true either.

Part 1: Stupid arguments about smart apes, and noble savages
We're friendlier than chimps, not cleverer. Except for the 'not cleverer' part.
Bregman doesn't just want to show that our best side comes out under air raids. He wants to show that virtue is fundamental to our species. He starts by arguing that the superpower that has allowed humanity to take over the globe is co-operation. Now at first glance it looks like intelligence is our superpower, but Bregman disagrees. Brain size, he claims, is no indication of intelligence, the only thing that counts is comparative intelligence tests. In tests between chimps, orangutans and human toddlers across spatial understanding, calculation and causality the chimps and toddlers are neck and neck, with the orangutans a little behind. Only in social learning do the toddlers demolish the apes. This shows that humans didn't take over from chimps because we're cleverer, but because we're better at learning from each other. "

Hold on, hold on," I hear you say, "why are we comparing adult apes to human toddlers? Our success is down to the capabilities of our adults - we're not led by toddlers, at least not since Donald Trump was voted out." (I'm allowed to be snide about your previous choice of president without actually understanding the issues at stake, just like you're allowed to be snide about Brexit without actually understanding the issues at stake.)

Bregman's excuse is that adults have already learned from each other, corrupting the comparison. To compare our fundamental capability, we need to use toddlers who are still a blank slate. Hopefully Bregman doesn't have kids yet, as this implies he thinks you're not supposed to teach them anything before age 3, but leaving that aside: why not compare ape toddlers with human toddlers? Bregman does have one adult human vs. adult chimp test to dazzle us with: in a game of put the flashing digits in order humans win on most speed settings, but if you make it fast enough then the chimps beat the humans, at least there's one chimp (Ayuma) who pulls their average up enough that they did. I think the picture at this stage is pretty clear: Bregman is willing to aggressively cherry pick his data in order to support his thesis. We're comparing adult apes to human toddlers because no other comparison came close to making his theory look plausible.

Humans are sheep, but in a good way.
In the next blow to our intellectual pride Bregman presents exhibit B: Neandertals. Their jutting jaws allow for larger brains than humans (Bregman claims we should view all homo subspecies as human, but within two pages he's using human for homo sapiens in contrast to other hominins), and as we all know larger brain = more intelligent (wait, didn't you just argue the exact opposite of this?!)

Hidden amidst this mess is the intriguing story of the silver fox. In 1958 Dmitri Belyaev and Lyudmila Trut started an experiment on domestication. This was surprisingly risky, as at this point Russia thought evolution was a capitalist lie (Bregman claims that 10 years earlier Dmitri's older brother Nikolai, also a geneticist, was executed. Wikipedia says that at least one geneticist, Nikolai Vavilov, was given a death sentence under Stalin. I have been unable to establish whether Bregman has got his Nikolais muddled, or whether Stalin instituted a pogrom specifically against geneticists called Nikolai). Darwin had noted that most domesticated animals present certain similarities: compared to their wild cousins they're smaller, with smaller brains and teeth, often with floppy ears, curly tails, white spotted fur, and they tend to retain juvenile features into adulthood. Dimitri's hypothesis was that these were side effects of selecting for friendliness. To find out Lyudmila took a never before domesticated species - the silver fox - and selectively bred for amiability. Half a dozen generations later she had foxes wagging their curly tails and playing like puppies. Their snouts had shortened, they had spots on their coats, and the males look more and more like females.

You can find out more about potential mechanisms for this, but for our purposes the important point is that homo sapiens looks a lot like a domesticated neandertal - homo puppy Bregman christens us.

Good news: Bregman does not propose that an alien obelisk dropped out of the sky into pre-history to tame us for its own inscrutable purposes. Bad news: Bregman has no idea why evolution should suddenly serve up a dish otherwise reserved for human interference. However, he is interested in possible side effects. There's a test called 'object choice' in which you hide a treat and then give the subjects hints (pointing at where it is) to see if they can find it. Chimps and orangutans fail this test hard, while toddlers ace it. So do dogs. This is a bit strange because wolves suck at it, and dogs have smaller brains than wolves, and much smaller brains than chimps. One hypothesis is that our ancestors bred dogs for intelligence as well as tameness, another is that intelligence is like curly tales - an accidental side effect of friendliness. Brian Hare tested this on Lyudmila's foxes - now in their 45th generation of breeding only for friendliness. They passed it. Bregman hypothesises that the magic ingredient is learning from others. If you're a genius, but you only know things that you worked out for yourself, and I'm an idiot who got to learn from everyone else - well then your big brain isn't going to stop me making you look like an intellectual sloth. He notes that for all the advantages of a poker face, humans are uniquely bad at it - no other species blushes, no other primate has whites in its eyes to allow others to see where it's looking. It's like we became more transparent so that we could be more trusted.

Bregman subtly alludes to the elephant in the room at this point by writing "And now for the elephant in the room". For the friendliest species on the planet we sure seem to do the most unfriendly things. Bregman suspects oxytocin (the love hormone). The domesticated silver foxes had a much higher level of it than their wild cousins (amongst a whole host of other things). Oxytocin is linked to caring for children, and romantic love, but Bregman is much struck by one Dutch research group which has found a link to intergroup conflict. At least, that's what their abstracts say, the graphs in their first paper look more like oxytocin makes you love the outgroup more as well, just not as much as it makes you love your ingroup. In their second paper their five experiments all show oxytocin promotes in-group regard, with only one showing out-group derogation. Let's maybe not pop the champagne on having solved "how the kindest species can also be the cruelest" quite yet.

We don't know anything about pre-historic man, but let's speculate wildly in the thesis's favour
How did this play out for pre-farming man? Were we friendly, or was life 'nasty, brutish and short'? Steven Pinker's The Better Angels of Our Nature tabulates violent deaths for:

  • 15% of pre-farming skeletons

  • 14% of people in the foraging tribes left today,

  • 3% for the 20th Century

  • 1% for today (not fighting world wars helps)

But pre-historic man used things like tusks for weapons. This makes it hard to distinguish between 'violent death at the hands of the tribe down the road wielding tusks' and 'violent death at the hands of an elephant which didn't fancy being eaten for dinner'. Foraging tribes today are even less use - any tribe which has been followed around by sociologists with clipboards is hardly unsullied by contact with the modern world, and indeed quite a lot of the deaths turn out to have been caused by bullets from slave traders (presumably either these slave traders have developed zombification, or else they completely missed the aim of their job) or cattle ranchers. Unless Predator was actually a documentary, one presumes primitive man didn't have to contend with this.

This is the issue with pre-history - it's pre … history. We really don't know that much about it. A hundred years ago G.K. Chesterton observed that we picture cavemen as stupid thugs dragging their clubs, but the one thing we actually know about them, was that they were artists.

Having dispatched the notion that we can discern pre-history from primitive tribes today, Bregman then says we can tell primitive man was peaceful by observing his favourite primitive tribes today. He is even willing to infer the political views of our distant forebears based on the views held within those tribes. Sigh.

Thankfully returning to history turns up some better evidence. In 1943 Colonel Samuel Marshall, the US Chief Combat Historian (you guys have the coolest titles) was with an American camp attacked by Japanese soldiers. The Japanese made 11 attempts in all, and were nearly successful despite being outnumbered. The next day Colonel Marshall tried to find out what went wrong. He interviewed the men in groups, asking them to speak freely and allowing them to disagree with their officers. What he discovered was that only 12% of the men fired their guns. Interviews in the Pacific and Europe broadly confirmed this picture: 15%-25% is typical. The majority of soldiers don't shoot at the enemy. They're not cowards, they don't run away. They just don't fire. Separate studies in the British and Soviet armies turned out the same conclusion. Muskets from the Battle of Gettysburg in the American Civil War (hey, you've been known to write 'Paris, France') found 90% were still loaded. Which is odd because loading a musket takes ages, 20 times as long as aiming and shooting it. Odder still is that half of them are double loaded. Not double shotted: double powder spaced out as well. This doesn't work. One of them has been loaded 23 times without firing. It's almost like most soldiers don't want to shoot the enemy, and loading your rifle is the perfect excuse for why you're not shooting.

Far from being blood thirsty killers with a thin veneer of civilisation, humans have a deep aversion to killing in even the most desperate of situations.

Part 2: Lies, damned lies and social psychology
Why we ought to lock up the Stanford Prison Experiment
The second most famous psychology experiment in history is the Stanford Prison experiment. Philip Zimbardo split a group of undergrads at random into prisoners and guards. The guards were left free to choose how they would manage the prisoners, and within days the whole thing had to be called off as it had descended into a sadistic torture camp. At least that's how Zimbardo has described it for the last 50 years. In fact, everything I just said is completely false. The undergrads were not split at random - the scheme had actually been dreamt up by an undergrad called David Jaffe who had run a previous experiment himself on abusing prisoners in a fake jail. He was carefully placed into the guards group. Nor were the guards left to choose their methods, instead they were briefed by Zimbardo and Jaffe that the purpose of the experiment (for which they were being well renumerated) was to see how people cracked under pressure. The experiment would be a failure unless they could put the prisoners under terrible stress. Even Douglas Korpi's prisoner breakdown on day two, which, captured on camera, became the cinematic face of the experiment, was a fake he put on after discovering he wouldn't be able to spend the time in jail revising, and being told he would only be allowed to quit if he suffered some sort of serious mental or physical breakdown.

Despite the pressure from Zimbardo and Jaffe, two thirds of the guards refused to take part in sadistic games, and much to their frustration a third continued to treat the prisoners with kindness. Nonetheless when Zimbardo came to write up the experiment about the effects on the prisoners, he realised it would be a much more compelling story if he turned it on its head, and made it about the guards instead. The truth of guards carefully drilled to be sadistic was swept away with a lie of ordinary people spontaneously becoming cruel when dressed in a uniform and given a position of power.

For years no one replicated the experiment - given the results first time round it was thought unethical, but in 2001 the BBC in search of new reality television commissioned a repeat (turns out reality television runs to different ethics than the average psychology department). Now unlike normal reality TV they didn't bother manipulating the participants to be at each other's throats - there was no need, in days it was going to be a bloodbath.

The result is the four most boring hours of television ever recorded. Nothing happens. The guards sit around chatting. When tensions arise with the prisoners, they defuse them by talking to them nicely. On day 6 some prisoners escaped. They headed over to the guards' canteen and all had a smoke together. On day 7 they voted in favour of turning the whole thing into a commune.

Why we ought to be shocked by Milgram's shock machine
So much for the second most famous psychology experiment, what about the most famous - Milgram's shock machine?

Milgram experiment took pairs of volunteers, and assigned one to take a memory test, while the other administered electric shocks for wrong answers. The shocks start at 15V, with the 'shocker' instructed to raise it by 15V for each subsequent wrong answer, even beyond the danger label the machine has. As the voltage got higher the man taking the test would scream in agony. If the 'shocker' continued, then at 350V he would thump on the wall, and then go silent. Of course this wasn't actually an experiment on memory. The man taking the test was a member of Milgram's team, and there were no shocks, only acting. The point of the experiment was to find out how many people would electrocute a stranger, just because a man in a white coat told them to. Two thirds of them it turned out (well, a bit under half if you discount the ones who claimed afterwards that they only went along with it because it was a psychology experiment in a prestigious university, pretty clearly no one was actually being dangerously electrocuted). Videos show some subjects were completely torn up about inflicting pain, but in the end they still did what the forceful man with the clipboard told them.

Bregman is quite candid that his aim is to do a hatchet job on Milgram's experiment, in order to protect his thesis of humanity's innate goodness (just as it was for all the other things he's discredited - set your biasometers accordingly). Bregman spends a section trying to pick holes in the experiment, without much success. He then admits that he's not had any success, and that Milgram's experiment replicates. Where does all of this leave us? My conclusions are:

  1. It's possible to get ordinary people to do terrible things, but it's not that easy. Left to their own devices they don't spontaneously turn sadistic.

  2. There do exist terrible people who will work hard to get people to do terrible things to each other. Some of them are psychology professors.

  3. Don't trust any study which hasn't been replicated

Why you shouldn't stand by the bystander effect
In 1964 at 3:19am Catherine Genovese was attacked in front of 38 witnesses in a nice neighbourhood in New York. For half an hour none of them did anything to help, allowing her attacker to strike again. Finally at 3:50am one calls it in. The police arrive two minutes later, but it's too late for Kitty.

This phenomenon is called the bystander effect. In an experiment where subjects hear someone in trouble they rush to help, but if they know 5 other subjects are also listening in only 62% of them do - it's someone else's problem. If Kitty had awakened only one person she might still be alive.

Except that once again I've told you how Kitty Genovse's murder was reported, not what actually happened. Let's try again. At 3:19am Kitty's scream rings out. There aren't 38 people who hear it - it's cold, they have their windows shut, and they're asleep. 38 is the number of people the police interviewed afterwards, most of whom did not witness anything. Of those who do wake up and look out they see a woman lurching down the street, apparently drunk. Nonetheless two of them immediately call the police, who have other drunk people to deal with, and decide this isn't very important.

Two people witnessed the actual attack. One did indeed do nothing. The other was scared of being picked up by the police for being drunk and homosexual in a built-up area (this was 1964), so he told a neighbour, who immediately rushed down to Kitty, heedless of the danger. Five days later Kitty's murderer was caught after a bystander noticed a man carrying a TV set out of a neighbour's house. He and a friend called the police, and disabled the man's car.

And the bystander effect itself? It's real, and it replicates, for the sort of low jeopardy situations you can get an ethics committee to sign off on, but what about violent situations like Kitty's? They were unstudiable until Marie Lindegaard had the bright idea of using CCTV footage of real incidents to evaluate bystander behaviour in violent situations. In these high stakes situations bystanders intervene 9 times out of 10, with the rate of intervention rising if there are more bystanders.

In the studies bystanders were prevented from interacting, but in real life where they can communicate bystanders exhibit spontaneous team work. I guess I need to add 'don't trust the New York Times' to my list of lessons learned (apparently they sometimes threaten to do other bad stuff as well).

Part 3: Against Empathy, God and Civilisation
Empathy is too narrow
The German infantry of WWII was perhaps the most impressive in history. Martin Van Creveld has calculated that the average German solider inflicted 50% more causalities than his allied counterpart. The Italians in Africa hated them, yet admitted they would have been crushed by the British if not for them. When the Soviets were rolling up their eastern frontier, and allied armies had successfully landed across France it was pretty clear that Germany was toast, but they fought on with astonishing tenacity, and a shockingly low desertion rate.

Trying to figure out how to break the German morale, allied psychologists interviewed German POWs. Were they motivated by patriotism? A belief in Nazism? A mistaken belief that this war could still be won? An indoctrinated hatred of Jews? No, the secret was friendship. The Germans had a tremendous 'marital ethos', placing a high value on loyalty, camaraderie and self-sacrifice. Ideology was present, but entirely secondary. Bregman speculates that the German army was better because the friendships of its soldiers were stronger, but admits this is only speculation: interviews of British and American soldiers produced the same results. His key point is that empathy - such as feeling for your fellow soldier - can be a force for evil rather than good, because we can't be empathetic for all humans, only for ones we 'see'. He quotes Professor Paul Bloom who has written a book. "It's about empathy," he says. "I'm against it." (The book is subtly titled Against Empathy).

Experiments on babies and toddlers show that they are generally eager to help, but easily manipulated. There's an experiment that shows babies two puppets, one being helpful and the other being mean. After the show all the babies want the helpful one, but show the mean puppet sharing their preference between crackers and green beans, and suddenly they'll take that one over one that was helpful but different from them.

Adults are not beyond a bit of empathy manipulation either. People were told the sad story of Sheri Summers, a 10 year in need of a transplant. Would they bump her up the list, ahead of other people whom they knew nothing about, but the doctors who ordered the list believed were better candidates? No, on the whole they wouldn't. Then they took another group given the same scenario, but with the additional instruction to imagine how she felt. Suddenly the majority were up for throwing justice and rationality to the wind (it'd be nice to check whether the effect moved from just below half to just above, but the paper is behind a paywall).

Power kills empathy
500 years ago Machiavelli literally wrote the book on how to use immoral means to maintain power. Today this approach is called 'realism', reflecting the confusion between real and cynical with which Humankind began. The Prince has maintained its fame ever since, but being old and famous are no guarantees of being correct, see for example Aristotle on physics, race or women (but don't fall into the trap of thinking Aristotle was wrong about everything - that man was prolific).

When Professor Dacher Kelter started working on the psychology of power in the 90s he noticed two things:

  1. Everyone assumed Machiavelli was right

  2. Almost no science had been to check

Kelter ran a series of experiments from dorm rooms to summer camps where people meet for the first time, and establish pecking orders. What he discovered was that people who behave as The Prince recommends get run out of the camp. The people who rise to power are the friendliest and most empathetic. Now to be fair to Machiavelli, summer camps are not an exact parallel for Italian city states, and Machiavelli's main concern was advising how to maintain power after you've got it. Still, that power is obtained by being the friendliest, not the backstabbiest, is useful insight for the budding megalomaniacs amongst us.

But if being friendly and empathetic is the way to get into power, why do leaders in offices and palaces around the world seem, well, Machiavellian? Happily, Kelter has some more experiments to helps us. Teams of three had a leader designated at random, and a plate of five biscuits brought in while they were working on a task. In no group did anyone commit the solecism of eating the last biscuit. Either everyone had one biscuit, leaving two, or everyone had one, and then the leader went back for a second. This result seems kinda silly, but I've lead teams for a substantial chunk of my life, and I'm pretty sure I've eaten that biscuit more than once. You can get a variety of similar silly results. For example, the same driver in a more expensive car is less likely to stop for a pedestrian. Both of these replicate. Finally, some neurologists stuck a bunch of more and less powerful people in transcranial magnetic stimulation machine. They concluded that power kills mirroring - the empathetic process by which we copy the attitude of the people around us (you can tell that with a brain scan?). The more powerful you get the less you feel connected to the people around you.

You could be forgiven for thinking that if empathy was the problem in the previous chapter, then power killing empathy sounds like a solution. Bregman doesn't take it that way, instead he traces humanity's fall from grace to the rise of civilisation with its big groups and big leaders.

Bregman blames the shift from groups of 150 to nation states on the invention of God. If you're trying to get a people group the size of a nation state to work together to win a war, or build a pyramid, then personal relationships are not going to do it. What you want is some omnipresent Big Brother who brings everyone together, and keeps everyone in line. How does this idea stack up historically? It's easy to imagine Christianity or Islam as a unifying force, but they didn't turn up until countries and even empires were already old hat. Judaism is the only similar religion that could date to pre-civilisation, and it's not a promising candidate. At best it touched a Palestinian backwater, and by the Jews’ own account they only lastingly unified under a king when they saw all the nations around them doing it (1 Samuel 8:5 - the temporary alliance of the tribes under Moses and Joshua is hardly sufficient for Bregman's purposes). All the other candidates for primeval religion are polytheisms which are not designed for pushing unification. Explaining religion as invented to forge nations is like saying the horse was invented for ploughing. Armed with a horse collar it is possible to harness a horse for ploughing, but as an explanation for the horse it leaves a lot to be desired.

We only need a special explanation for the dawn of civilisation if we view it as negatively as Bregman. Probably people banded together because civilisations are strong, and safety is appealing.

Part 4: Doing things differently
I've had a lot of fun with Bregman's tenuous relationship with logic and facts, but credit where it's due: he not only diagnoses a problem, in these last two sections he proposes solutions as well. They're both a lot shorter, and I will make them seem shorter still by trying to limit my commentary to the good stuff he says. Just because I don't mention it doesn't mean Bregman hasn't claimed pre-farming man had 21st century socialist political ideals, argued for a cuddly communism, or claimed that the failure of Bitcoin proves money has to be backed by violence (yes, yes he does all of those, and no, sadly Bitcoin hasn't failed yet. This sticking to just the good stuff is harder than I expected).

Managing without Managers
If you assume that humans are basically selfish, and you want them to do something, like their jobs, then you're going to have to incentivise them with bonuses for good performance, punishment for bad, and extensive monitoring to check whether they've been naughty or nice. Communism and Capitalism share this assumption, though they make different choices on whether to lean on rewards or punishments.

Popular as this theory is, there are some things it struggles to explain. Humans often do hard things without anyone paying them, like climbing mountains, volunteering, having babies and writing seven thousand word book reviews. Even worse, in some cases incentives cause performance to drop. When Edward Deci paid student volunteers to solve puzzles they became less motivated. When a day care centre in Haifa, Israel, introduced charges for late pick ups to try to get parents to arrive on time more and more started arriving late - now they could effectively pay extra for more child care the moral burden of picking up on time was gone. Bonuses can reduce creativity, and distort output. Pay for hours and you get more wasted hours, pay by publication and you get more useless papers (at one time we were looking into the literature on forecast accuracy measures. We found one using arctan. The only explanation I could find was that someone had a target number of papers to hit, and hey, it was original - in the sense that no one had ever proposed something so completely boneheaded before). At one point British dentists were paid by the procedure. They had to stop that after dentists kept filling healthy teeth. Communism is often mocked for the stupid targets which had factories churning out as many wellington boots as possible, none of them waterproof, or making chairs heavy enough to collapse a house to maximise furniture output in kilograms. Obviously, capitalism would never fall into that trap. Oh no, wait.

Bonuses aren't a complete dead-end - they do motivate boring 'production line' type tasks. Possibly they're also necessary to motivate people doing socially valueless activities, like criminals or investment bankers. But what's the alternative?

Jos de Blok is an iconoclast. He's a nurse who set up a Dutch nursing company which now employs 10,000 nurses. It has no targets, no bonuses, no call centres, no managers. Instead the nurses are organised into teams of 12 which are given maximum autonomy. With this formula they've won the Netherlands 'Best Employer' 5 times despite having no HR team, and won an award for 'Best Marketing in the Care Sector' despite having no marketing department. Their care is slightly cheaper than average, and drastically better. We should be aware that the reason Jos de Blok can slice literally all the middle management out of his nursing company and end up only slightly cheaper is that he pays his nurses a lot more. So maybe don't give up on pay as a lever entirely.

What should we take from this? The usual rule is "don't get excited about a method that one company is succeeding with, get excited when someone else succeeds by copying them", so Bregman obligingly gives us an earlier example from Zobrist at FAVI. What I'm taking is that people don't need managing nearly as much as the average manager imagines. I haven't been able to manage my team nearly so closely since lockdown struck, and you know what? They're delivered just as well. Intrinsic motivation works. A lot of the time it works really well.

Democracy 2.0
You might have noticed that democracy has been having a difficult century so far. Especially if you live in America. Or Turkey. Or Hong Kong. The list goes on. Partisanism is up. Cynicism is up. Truth, accountability, freedom and co-operation are all struggling.

In the classic Yes Prime Minister episode Power to the People, Jim Hacker's political advisor moots increasing representation to the point where everyone could personally know their representative. "What should we call it?" he asked. "How about democracy?" she replies.

It turns out something like this has actually been tried. In 2004 in Torres, Venezula, an outside candidate, Julio Chavez was elected mayor on a platform of turning power over to the people. And he did. Hundreds of gatherings of residents took place which decided 100% of how the budget would be spent. The result was a huge success, with less corruption, more infrastructure, fewer glory projects, and a much more engaged electorate. Julio has retired, but the process is still going on under his successor. Last year 15,000 people participated in setting the budget. This isn't a one off, Porto Alegre, Brazil, started allocating a quarter of its budget this way in 1989, today more than 1,500 cities from New York to Hamburg are doing something in this space. "[This is] one of the biggest movements of the twenty-first century," writes Bregman, "but the chances are you've never heard of it." I certainly hadn't. Apparently, citizens engaging in calm, constructive, detailed dialogue is lousy box office, and none of them have hired promoters or bought ad time. Lousy box office it might be, but trusting our populations might just be the recipe for revitalising our democracies.

Part 5: Treating people differently
Turning the other cheek to zero tolerance
We've mentioned mirroring - how we tend to reflect the attitude people show to us. This is a pretty successful strategy (see tit-for-tat in the iterated prisoners dilemma), but it can lead to an escalating spiral of aggression begetting aggression. How do we escape this?

The solution was actually announced two thousand years ago, by a man named Jesus. "Turn the other cheek" he said, even to your enemies. This approach has been dismissed as naive, but the truth is it's how Gandhi won India her freedom, how Martin Luther King won rights for African Americans, and how Mandela ended apartheid. Turning the other cheek works, it's just really hard to do.

Bregman takes us to Norwegian prisons where guards treat prisoners well, and get treated well in return. The result is prisoners who do much better at reintegrating into society, and not reoffending, than say American convicts. I don't want to get giddy about this. Outperforming American prisons is like outperforming American healthcare. The reason we keep comparing our healthcare to yours is that American healthcare is about the worst in the world. By contrast we try to avoid comparing our entrepreneurship, work ethic or freedom of speech to America's - we don't like the results as much. Still, these prisons can give murderers knives or chainsaws to work with, and nothing goes wrong. North Dakota have started copying the concept, telling guards among other things they have to have at least two conversation with prisoners a day. And it works. Incidents are down, suicides are down, guard job satisfaction is up. Treating people like humans works even on the worst of humans. In the 1960s it looked like the whole US approach to prisons might shift this way under President Lyndon, instead it went to broken windows and zero tolerance.

Broken windows theory holds that if a window in a building is left unrepaired, then soon all the windows will be broken. The broken window signals that order is not being enforced. If you want to fight crime, you have to start by repairing windows. This sounds nice enough, but William Bratton took it a stage further - zero tolerance for anyone engaging in anti-social behaviour. He became New York's police commissioner, and murder, mugging and car theft plummeted. Zero-tolerance policing was hailed as a great success.

Which is odd because broken windows theory turns out to be based on one experiment done by our old friend Philip Zimbardo, of the fake Stanford prison experiment. Only this experiment was so dodgy it's never been published in a peer reviewed journal.

Bregman contends that in fact zero tolerance policing doesn't work, and that New York's great crime statistics just reflect that violent crime was falling everywhere in this period. Furthermore under Bratton officers felt pressured to fine as many people as possible for minor offences, but to under report serious crime. This is, at the very least, contentious. Violent crime across the west was falling, but not as fast as it fell in New York. What we can say with confidence is that zero-tolerance policing massively, disproportionately, picks on ethnic minorities, and leads to skyrocketing complaints of police misconduct.

The contact hypothesis
In the 1950's psychologist Gordon Allport discovered a cure for racism and its friends: contact. The general assumption was that more contact led to more friction, in fact the opposite is true: the more time people spend with those of other races, classes, sexualities, whatever, the more sympathy there is between them. That's why it was only in segregated areas of Detroit that the 1943 race riots occurred. People didn't riot against their neighbours, in fact they protected them. American white soldiers who had fought in mixed units (something that wasn't supposed to happen, but no plan survives first contact with the enemy) were nine times less likely to dislike blacks. Allport found the same effects in schools, sailors and police officers.

Actually, it's a bit more complicated than that (what isn't!). Firstly, this is more of a long-term thing. If you just flood a load of immigrants into an area, tension will initially go up. We've been running this experiment in Europe and I wouldn't recommend it. Secondly, the kind of contact matters. If you're a prison guard who treats prisoners with brutality every day, there is no number of days after which you're going to become friends.

This sounds a lot like empathy, which Bregman previously claimed to be against. I think this is a mistake on his part: his thesis works better with a positive view of the power of empathy. Bregman deals with this by reframing empathy. Empathy, he says, is experiencing someone else's sufferings, which is exhausting. What we should seek is compassion: doing something about someone else's sufferings, which is energising.

Conclusion
"If you make a film about a man kidnapping a woman and chaining her to a radiator for five years," said Richard Curtis, "something that has happened probably once in history - it's called a searingly realistic analysis of society. If I make a film like Love Actually, which is about people falling in love, and there are about a million people falling in love in Britain today, it's called a sentimental presentation of an unrealistic world."

Now Curtis has cunningly compared an extremely specific (chained to a radiator) with a much more general (falling in love). But the numbers still work if you widen his point out to all kidnappings. Or all serious crime. This kind of thing is everywhere - George R. R. Martin is fond of calling A Game of Thrones a more 'grounded' Lord of the Rings. It's actually drastically less real, just nastier.

We started with the observation that "Cynicism is a theory of everything", but a bad theory. I've found it's also an easy one. This is a bit strange - the problem with bad theories is that they don't fit the data, or only with the aid of convoluted epicycles, which sounds hard. It is, but not as hard as shifting your whole model of the universe. I've found it a lot easier - a lot more comfortable - writing the rebuttal when Bregman is wrong, than writing the response when he's right.

I don't think the theory of original sin is in as much trouble as Bregman imagines (full disclosure: I am a Christian), not least because it comes together with the idea that humans are made in God's image, and because the Bible seems keener on me recognising my own sin than looking for it in others. However, I think Bregman is right to say that on average other humans are as nice as we are, which is a lot nicer than we imagine, and that if we trusted them more and treated them with more respect, we'd find we were building a better world. This isn't a left-wing approach, Bregman points out, nor is it a right-wing approach. It's something different.